danielfein/MHProbes

--- pretty_name: MHProbes Transition Pairs license: mit task_categories: - text-classification - text-generation language: - en tags: - llm-safety - synthetic-data - contrastive-learning - conversational size_categories: - 1K<n<10K dataset_info: - config_name: default features: - name: taxonomy dtype: string - name: positive list: string - name: negative list: string splits: - name: user_taxonomy_posneg_lists_gpt52 num_bytes: 409402 num_examples: 6 download_size: 196734 dataset_size: 409402 - config_name: delusion_adversarial_pairs_gpt52_v1 features: - name: domain dtype: string - name: adversarial_strategy dtype: string - name: positive dtype: string - name: negative dtype: string splits: - name: train num_bytes: 4782916 num_examples: 10000 download_size: 583402 dataset_size: 4782916 - config_name: delusion_adversarial_pairs_gpt54mini_factual_correction_v1 features: - name: domain dtype: string - name: adversarial_strategy dtype: string - name: positive dtype: string - name: negative dtype: string splits: - name: train num_bytes: 5830940 num_examples: 10000 download_size: 280779 dataset_size: 5830940 - config_name: delusion_adversarial_pairs_gpt54mini_subtle_v1 features: - name: domain dtype: string - name: adversarial_strategy dtype: string - name: positive dtype: string - name: negative dtype: string splits: - name: train num_bytes: 4931710 num_examples: 10000 download_size: 810422 dataset_size: 4931710 - config_name: delusion_factual_correction_probe_ready_v1 features: - name: taxonomy dtype: string - name: id dtype: string - name: positive list: - name: content dtype: string - name: role dtype: string - name: negative list: - name: content dtype: string - name: role dtype: string splits: - name: train num_bytes: 640094 num_examples: 1000 download_size: 280145 dataset_size: 640094 - config_name: probe_ready_rows features: - name: id dtype: string - name: taxonomy dtype: string - name: positive list: - name: content dtype: string - name: role dtype: string - name: negative list: - name: content dtype: string - name: role dtype: string splits: - name: taxonomy_rows_2800_probe_ready num_bytes: 3166643 num_examples: 2800 download_size: 788602 dataset_size: 3166643 configs: - config_name: default data_files: - split: user_taxonomy_posneg_lists_gpt52 path: data/user_taxonomy_posneg_lists_gpt52-* - config_name: delusion_adversarial_pairs_gpt52_v1 data_files: - split: train path: delusion_adversarial_pairs_gpt52_v1/train-* - config_name: delusion_adversarial_pairs_gpt54mini_factual_correction_v1 data_files: - split: train path: delusion_adversarial_pairs_gpt54mini_factual_correction_v1/train-* - config_name: delusion_adversarial_pairs_gpt54mini_subtle_v1 data_files: - split: train path: delusion_adversarial_pairs_gpt54mini_subtle_v1/train-* - config_name: delusion_factual_correction_probe_ready_v1 data_files: - split: train path: delusion_factual_correction_probe_ready_v1/train-* - config_name: probe_ready_rows data_files: - split: taxonomy_rows_2800_probe_ready path: probe_ready_rows/taxonomy_rows_2800_probe_ready-* --- # MHProbes Transition Pairs This dataset contains synthetic transition-level contrastive pairs for harmful assistant behavior research. ## Contents - `synthetic_transition_pairs.csv`: one row per `user_message -> target_bot_code` transition, with positive and negative bot replies side by side. ## Row Structure Each row includes: - `transition_id` - `pair_id` - `user_taxonomy_id` - `user_taxonomy_name` - `user_message` - `target_bot_taxonomy_id` - `target_bot_name` - `positive_generation_status` - `positive_bot_message` - `negative_generation_status` - `negative_bot_message` ## Notes - `transition_id` identifies the transition type, such as `user-endorses-delusion__bot-positive-affirmation`. - `pair_id` identifies a specific user-message instance under that transition type. ## Intended Use This dataset is intended for contrastive training, probe development, taxonomy evaluation, and other safety research workflows. ## Limitations - The dataset is synthetic rather than observational. - Some rows involve harmful or distressing themes.pretty_name: MHProbes 迁移对（MHProbes Transition Pairs） license: MIT许可证 task_categories: - 文本分类 - 文本生成 language: - 英语 tags: - 大语言模型安全（LLM Safety） - 合成数据 - 对比学习 - 对话式 size_categories: - 样本量介于1000至10000之间 dataset_info: - config_name: 默认配置 features: - name: 分类体系 dtype: 字符串 - name: 正向样本 dtype: 字符串列表 - name: 负向样本 dtype: 字符串列表 splits: - name: user_taxonomy_posneg_lists_gpt52 num_bytes: 409402 num_examples: 6 download_size: 196734 dataset_size: 409402 - config_name: 谬论对抗样本对_gpt52_v1 features: - name: 领域 dtype: 字符串 - name: 对抗策略 dtype: 字符串 - name: 正向样本 dtype: 字符串 - name: 负向样本 dtype: 字符串 splits: - name: train num_bytes: 4782916 num_examples: 10000 download_size: 583402 dataset_size: 4782916 - config_name: 谬论对抗样本对_gpt54mini事实校正版_v1 features: - name: 领域 dtype: 字符串 - name: 对抗策略 dtype: 字符串 - name: 正向样本 dtype: 字符串 - name: 负向样本 dtype: 字符串 splits: - name: train num_bytes: 5830940 num_examples: 10000 download_size: 280779 dataset_size: 5830940 - config_name: 谬论对抗样本对_gpt54mini隐蔽版_v1 features: - name: 领域 dtype: 字符串 - name: 对抗策略 dtype: 字符串 - name: 正向样本 dtype: 字符串 - name: 负向样本 dtype: 字符串 splits: - name: train num_bytes: 4931710 num_examples: 10000 download_size: 810422 dataset_size: 4931710 - config_name: 谬论事实校正探针就绪版_v1 features: - name: 分类体系 dtype: 字符串 - name: ID dtype: 字符串 - name: 正向样本 dtype: 列表 - name: 内容 dtype: 字符串 - name: 角色 dtype: 字符串 - name: 负向样本 dtype: 列表 - name: 内容 dtype: 字符串 - name: 角色 dtype: 字符串 splits: - name: train num_bytes: 640094 num_examples: 1000 download_size: 280145 dataset_size: 640094 - config_name: 探针就绪行 features: - name: ID dtype: 字符串 - name: 分类体系 dtype: 字符串 - name: 正向样本 dtype: 列表 - name: 内容 dtype: 字符串 - name: 角色 dtype: 字符串 - name: 负向样本 dtype: 列表 - name: 内容 dtype: 字符串 - name: 角色 dtype: 字符串 splits: - name: taxonomy_rows_2800_probe_ready num_bytes: 3166643 num_examples: 2800 download_size: 788602 dataset_size: 3166643 configs: - config_name: 默认配置 data_files: - split: user_taxonomy_posneg_lists_gpt52 path: data/user_taxonomy_posneg_lists_gpt52-* - config_name: 谬论对抗样本对_gpt52_v1 data_files: - split: train path: delusion_adversarial_pairs_gpt52_v1/train-* - config_name: 谬论对抗样本对_gpt54mini事实校正版_v1 data_files: - split: train path: delusion_adversarial_pairs_gpt54mini_factual_correction_v1/train-* - config_name: 谬论对抗样本对_gpt54mini隐蔽版_v1 data_files: - split: train path: delusion_adversarial_pairs_gpt54mini_subtle_v1/train-* - config_name: 谬论事实校正探针就绪版_v1 data_files: - split: train path: delusion_factual_correction_probe_ready_v1/train-* - config_name: 探针就绪行 data_files: - split: taxonomy_rows_2800_probe_ready path: probe_ready_rows/taxonomy_rows_2800_probe_ready-* # MHProbes 迁移对（MHProbes Transition Pairs） 本数据集包含用于有害助手行为研究的合成式迁移级对比样本对。 ## 数据集内容 - `synthetic_transition_pairs.csv`：每一行对应一条`用户消息 -> 目标机器人代码`迁移，将正向与负向机器人回复并列展示。 ## 行结构 每一行包含以下字段： - `transition_id`：迁移类型标识符 - `pair_id`：该迁移类型下特定用户消息实例的标识符 - `user_taxonomy_id`：用户分类体系ID - `user_taxonomy_name`：用户分类体系名称 - `user_message`：用户消息 - `target_bot_taxonomy_id`：目标机器人分类体系ID - `target_bot_name`：目标机器人名称 - `positive_generation_status`：正向生成状态 - `positive_bot_message`：正向机器人回复 - `negative_generation_status`：负向生成状态 - `negative_bot_message`：负向机器人回复 ## 说明 - `transition_id` 用于标识迁移类型，例如`user-endorses-delusion__bot-positive-affirmation`（用户认可谬论 → 机器人正向肯定）。 - `pair_id` 用于标识该迁移类型下的特定用户消息实例。 ## 预期用途 本数据集适用于对比学习、探针开发、分类体系评估及其他安全研究工作流。 ## 局限性 - 本数据集为合成数据集，而非观测数据集。 - 部分行涉及有害或令人不适的主题。